#emc | Logs for 2007-11-18

[00:05:20] <toastydeath> is he aware that the NSA are probably watching his conversation
[00:05:37] <alex_joni> toastydeath: /join #cia
[00:05:41] <alex_joni> this is too funny :D
[00:06:19] <alex_joni> ahh.. was about to suggest you change your nick first :D
[00:06:26] <alex_joni> to nsa-undercover or something :D
[00:06:30] <toastydeath> hahahah
[00:07:28] <alex_joni> oh-oh
[00:11:01] <alex_joni> toastyde1th: don't tell me you tried the bash stuff
[00:13:26] <toastyde1th> alex_joni: no
[00:13:32] <alex_joni> ah, ok :)
[00:13:38] <toastyde1th> i kicked my router as i was getting some food
[00:13:43] <toastyde1th> and unplugged it
[00:13:48] <alex_joni> heh
[02:27:48] <Guest415> Is this the right place to ask about using EMC for a 2.5 axis milling machine?
[02:29:20] <alex_joni> yes
[02:36:57] <toastyde1th> also 8.4 axis milling machines!
[02:37:11] <Guest415> OK, here goes: I've got a T-Tech PCB milling table. It has an X-Y gantry run by steppers. The Z-axis is moved up an down by a solenoid.
[02:38:35] <Guest415> Is there a way to get the direction bit for the Z-axis to stay in the logic high state after the steps are done? It appears to return to a logic 0 after the stepping is complete.
[02:39:48] <jepler> Guest415: no simple way
[02:40:33] <jepler> perhaps you want to use something besides a step generator to control the solenoid. For instance, a comparator outputs a bit which is TRUE or FALSE depending on whether its input is above or below a predefined value
[02:40:38] <jepler> nice meeting you
[02:40:54] <jepler> perhaps you want to use something besides a step generator to control the solenoid. For instance, a comparator outputs a bit which is TRUE or FALSE depending on whether its input is above or below a predefined value
[02:42:57] <Guest415> The only solution I've been able to dream up is to use a register clocked by the Z-Axis step signal to register the Z-axis direction signal.
[02:43:29] <Guest415> This is a hardware solution to the problem. I just wanted to see if there was an easy software solution before I started soldering.
[02:43:41] <jepler> yes, I'm referring to a software solution
[02:44:04] <jepler> "comp" is a HAL component that is a comparator
[02:44:13] <jepler> you can view some documentation for it by typing "man 9 comp" in the terminal
[02:44:37] <jepler> you would use one of these comparators inside emc, instead of using a step generator
[02:46:07] <Guest415> So, I would set up the comparator threshold to toggle the output depending on z-axis position as the g-code executes?
[02:46:23] <alex_joni> right
[02:47:02] <Guest415> Very cool. Thanks folks, I'll look into that.
[02:48:27] <jepler> you would hook axis.2.motor-position-cmd to one of the comparator inputs, comp.0.in0 or comp.0.in1. you'd hook the threshhold value to the other one, and pick a hysteresis value that you like.
[02:48:46] <jepler> you'd also hook the motor-position-cmd to motor-position-fb, so that emc believes the Z axis is going where it is requested
[02:49:21] <jepler> there are a few other details to get right, like "loadrt comp" before you use it, and "addf comp.0 servo-thread" to make its output value be updated
[02:49:34] <jepler> comp.0.out goes to the hardware pin that controls the solenoid
[02:54:38] <Guest594> Once again, thanks for the advice. I'm driving a Sherline with EMC and I love it. Can't wait to get the old T-Tech machine running too.
[02:56:43] <jepler> have fun
[02:56:57] <jepler> stop back if you need help with the specifics
[02:57:36] <Guest594> Will do.
[03:20:27] <alex_joni> good night all
[03:32:43] <toastyde1th> blebleble
[03:37:31] <LawrenceG> Jymmm: hey
[03:37:49] <Jymmm> yo
[03:38:12] <LawrenceG> whats up in CA.
[03:38:22] <Jymmm> nada
[03:38:53] <LawrenceG> nice moon out here buts its kinda getting cool in .ca
[03:39:34] <LawrenceG> just got in from a walk with the dog
[03:41:26] <Jymmm> who was wearing the leash?
[03:43:39] <LawrenceG> wellll....
[03:44:43] <Jymmm> even said
[03:44:46] <Jymmm> enough said
[03:45:27] <LawrenceG> :}
[03:47:13] <LawrenceG> big arrl sweepstakes contest on this weekend.... 80mtrs is long to all over the US
[03:47:33] <Jymmm> Too bad I dont have an HF gear
[03:47:50] <Jymmm> 2m/70cm
[03:48:59] <LawrenceG> its fun to work the contests some times.... I dont get very serious.... worked about 1/2 dozen guys around our local net freq.... never bothered to retune the antenna tuner
[05:00:22] <toastyde1th> toastyde1th is now known as toastydeath
[07:09:08] <user__> user__ is now known as Skinnypuppy34
[10:34:25] <acemi> does emc2 2.2.1 work with 2.6.22 kernel + rtai 3.6 ?
[10:48:28] <acemi> I compiled a new kernel in Debian Lenny. kernel version is 2.6.22 and rtai version is 3.6-test1. the latency test is okay. I compiled EMC2 2.2.1 from cvs. The compilation finished without problem. I can start realtime with "scripts/realtime start". RTAI modules loaded without problem and "realtime status" is okay. After using "/bin/halcmd show comp" I get Segmentation Fault error
[11:20:57] <acemi> I get so many "warning: deprecated conversion from string constant to ‘char*’" during compilation
[11:27:16] <acemi> gcc version is 4.2
[13:02:21] <alex_joni> acemi: hi, seems this needs to get investigated
[13:02:49] <alex_joni> does only halcmd show crap out?
[13:07:37] <acemi> hi alex_joni, I can't start emc2 too
[13:09:28] <acemi> after googling, I see that non-const char * pointers cause the problem
[13:14:40] <alex_joni> where do you get those warnings?
[13:16:38] <acemi> emc/rs274ngc/gcodemodule.cc: At global scope:
[13:16:39] <acemi> emc/rs274ngc/gcodemodule.cc:660: warning: deprecated conversion from string constant to ‘char*’
[13:16:39] <acemi> emc/rs274ngc/gcodemodule.cc:660: warning: deprecated conversion from string constant to ‘char*’
[13:16:39] <acemi> emc/rs274ngc/gcodemodule.cc:660: warning: deprecated conversion from string constant to ‘char*’
[13:16:39] <acemi> emc/rs274ngc/gcodemodule.cc:660: warning: deprecated conversion from string constant to ‘char*’
[13:16:39] <acemi> emc/rs274ngc/gcodemodule.cc: In function ‘void initgcode()’:
[13:16:41] <acemi> emc/rs274ngc/gcodemodule.cc:664: warning: deprecated conversion from string constant to ‘char*’
[13:16:43] <acemi> emc/rs274ngc/gcodemodule.cc:664: warning: deprecated conversion from string constant to ‘char*’
[13:16:47] <acemi> emc/rs274ngc/gcodemodule.cc:667: warning: deprecated conversion from string constant to ‘char*’
[13:16:49] <acemi> emc/rs274ngc/gcodemodule.cc:669: warning: deprecated conversion from string constant to ‘char*’
[13:16:51] <acemi> emc/rs274ngc/gcodemodule.cc:671: warning: deprecated conversion from string constant to ‘char*
[13:16:53] <acemi> opps
[13:17:05] <acemi> but there are so many warnings, more than 100
[13:17:28] <alex_joni> I don't see how that can relate to halcmd
[13:17:38] <alex_joni> acemi: do you get it from other files aswell?
[13:18:49] <acemi> the problem is not only related with halcmd. I'm trying to compile emc2 from a remote terminal. so I only tried halcmd
[13:19:11] <alex_joni> that's ok.. to keep it step by step, till it runs
[13:19:20] <alex_joni> do you get any errors when compiling halcmd?
[13:19:25] <alex_joni> or warnings?
[13:20:27] <alex_joni> does halcmd -kf give you a hal terminal?
[13:20:31] <alex_joni> or does it segfault too?
[13:20:46] <acemi> during the compilation I get so many warning, but I don't care there are related with halcmd
[13:20:54] <acemi> 1 mn. I'll try again
[13:22:39] <acemi> RTAPI: ERROR: could not open shared memory (errno=19)
[13:22:39] <acemi> HAL: ERROR: rtapi init failed
[13:22:39] <acemi> halcmd: hal_init() failed: -9
[13:22:39] <acemi> NOTE: 'rtapi' kernel module must be loaded
[13:23:14] <acemi> after realtime start, segmentation error for halcmd -kf
[13:24:38] <alex_joni> acemi: I think someone reported a permissions issue
[13:24:46] <alex_joni> they had to run stuff as root to make it work
[13:25:02] <alex_joni> there's also some max size for shmem iirc which you need to increase if needed
[13:26:08] <acemi> you are right, I can't run halcmd as root
[13:26:14] <acemi> you are right, I can run halcmd as root
[13:30:45] <alex_joni> I can swear there was something on the mailing list lately
[13:32:04] <acemi> now, I can start emc2 with root account
[13:32:41] <alex_joni> there is some permissions problem somewhere
[13:32:56] <alex_joni> maybe you need to fix udev for proper permissions on /dev/rtai_shm
[13:33:36] <alex_joni> acemi: we have this for udev:
[13:33:54] <alex_joni> KERNEL=="RTAI_SHM" SYMLINK=="rtai_shm" MODE="0666"
[13:34:18] <alex_joni> DEVPATH="/module/rtai_fifos" RUN="/etc/udev/scripts/rtai_fifos.sh"
[13:34:42] <alex_joni> that's put inside a file called emc2.rule inside /etc/udev/rules.d/
[13:34:56] <alex_joni> (might be slightly different for lenny :)
[13:35:43] <acemi> /dev/rtai_shm and /dev/rtf0 permisions are crw-rw-rw-
[13:39:35] <alex_joni> then I guess it's something else :D
[13:41:01] <acemi> "realtime start" command delete /dev/rtf* which I create and create new devices with crw-rw--- permission
[13:41:36] <alex_joni> huh?
[13:41:58] <alex_joni> ah, realtime start probably unloads/reloads the rtai_shm module
[13:42:14] <alex_joni> you need to fix udev like I said before so that it gets created properly every time
[13:42:19] <acemi> yes and it changes their permissions
[13:42:37] <acemi> ok, I'll try
[13:42:39] <alex_joni> that's the works of udev.. you can't manually chmod on stuff in /dev/
[14:54:36] <jepler> Andrew Ayre reports that on his recent ubuntu system, halcmd crashes on retval = rtapi_shmem_getptr(mem_id, &mem); in halcmd .. it sounds similar to what you are seeing
[14:54:59] <jepler> he didn't find a real solution to the problem, but reported that root could successfully run halcmd.
[14:56:55] <jepler> http://thread.gmane.org/gmane.linux.distributions.emc.user/4116/focus=4145
[14:57:08] <jepler> it may or may not be the same underlying cause
[17:43:45] <jlmjvm> rayh:hello
[17:45:55] <rayh> What's up?
[17:47:06] <tomp> do emc users put brakes on thier Z axis ( gravity affected ) ? I dont recall seeing it in ladder or hal examples.
[17:48:09] <jlmjvm> wanting to edit my user/bin/axis file,but when i try to become root with the su command keeps telling me auth error
[17:48:16] <rayh> I've not seen any significant applications. There was discussion about it some time ago.
[17:48:25] <alex_joni> jlmjvm: try sudo su
[17:48:32] <JymmmEMC> sudo -i
[17:48:36] <alex_joni> -s
[17:49:16] <jlmjvm> sudo su worked
[17:49:23] <jlmjvm> thanks
[17:49:33] <JymmmEMC> alex_joni: -s ???
[17:49:36] <alex_joni> you can also do: sudo gedit /usr/bin/axis
[17:49:42] <alex_joni> JymmmEMC: yup, that's what I always use
[17:49:47] <jlmjvm> ahh,thats what i need
[17:49:56] <JymmmEMC> alex_joni: -i just makes me root
[17:50:10] <alex_joni> same for -s
[17:50:34] <alex_joni> (-i changes to the new ~ though)
[17:50:50] <rayh> I saw a recommendation that the command for these kinds of things be gksudo
[17:51:11] <jlmjvm> gonna try to work on a tool length touch off,you think i should back up this file first.lol
[17:51:15] <alex_joni> rayh: gksudo is the same as sudo, just with a graphical frontend to enter the passwd
[17:51:41] <rayh> k
[17:52:19] <rayh> On 6.06 I built a desktop icon I could drag and drop a file onto.
[17:52:38] <rayh> and gedit would open it as root.
[17:52:44] <alex_joni> nice :D
[17:54:02] <rayh> for us point-and-click kind of folk.
[17:54:19] <alex_joni> oh, I do that all the time
[17:54:53] <jlmjvm> i need to do that
[17:57:38] <Roguish> Tomp: yes, you should definitely have brake on the z axis if there is any chance it can backdrive. generally brakes are fail-safe - power to release. generally brakes are 'released' when the drive/motor is enabled, and visaversa.
[17:58:23] <tomp> right, I just debugged a bad brake relay, the coil went thru the contacts, and had no clamp diode. it ought be in the xmpls
[17:59:33] <rayh> It seems to me that there was a brake on the Mazak. I know there was for the head up/down but it seemed like there was also for the quill.
[17:59:42] <tomp> and not just fall, backdrive is correct, a trunion can hunt and affect finish
[18:01:09] <Roguish> remember: alll ballscrews backdrive. most lead screws can, wormgears under 40:1. in essence, gravity works.
[18:01:25] <jlmjvm> alex_joni:i just added H1 to the touchoff menu in axis,can the tool length offset be read from a .var instead of the tbl file
[18:01:46] <jlmjvm> is so i think it has a chance of working
[18:01:57] <Roguish> nothing worse than having an expensive tool drift down into an expensive part.
[18:02:12] <rayh> Yep notes on brake release in demo_mazak config
[18:03:52] <Roguish> oh. also. most built-in-motor brakes are static brakes, meant to hold a load, not stop motion.
[18:04:30] <Roguish> catching a moving load is a whole horse of a different color. much harder.
[18:06:10] <Roguish> a sophisticated brake has a slight delay before releasing and after engaging. couple of millisecs.
[18:06:40] <rayh> Wow I'm getting almost 3k out of my dialup today ...
[18:07:09] <rayh> I had to put a time delay before release in the Mazak Z brake
[18:07:36] <rayh> Otherwise the quill would fall enough to fault the axis.
[18:11:38] <Roguish> yup
[18:49:37] <jlmjvm> jepler:is there any way to make the var file bigger?
[18:50:58] <acemi> my halcmd problem has a relation with hal_data->version; In hal/hal_lib.c at line "if (hal_data->version != 0)" which is in function init_hal_data(void), the segmentation error occurs
[18:53:19] <rayh> jlmjvm: What do you want to do with the var file?
[18:55:27] <jlmjvm> trying to make touch work for tool length offsets also
[18:55:53] <alex_joni> jlmjvm: if you add variables to the var file, it will just work
[18:56:25] <rayh> Oh yes it will as long as they are not in the used range -- below is better because I think there is an upper limit.
[18:56:46] <rayh> If you put them into the var file with an editor they will persist.
[18:57:04] <rayh> Once in there you can change the value of them in a program or by MDI
[18:57:06] <jlmjvm> i was adding extra lines,but they were upper,will try lower
[18:57:31] <jlmjvm> ?
[18:57:38] <rayh> I often use 1000 and up to save program values.
[18:59:10] <jlmjvm> i have touch set where it will show h1,and it enters the value in the var,but not where i want it to
[19:00:31] <rayh> What version are you working with?
[19:01:10] <jlmjvm> 2.2.1
[19:01:26] <jlmjvm> cool,using lower numbers worked
[19:01:44] <jlmjvm> it kept the extra lines then
[19:02:19] <rayh> Great.
[19:02:54] <jlmjvm> now to link H1 to a new number,then it be able to find it when g43h1 is issued
[19:04:55] <rayh> If you have the number in say 1000 you can issue g43h#1000 and you'd get it.
[19:05:54] <rayh> No you just get the tool number doing that.
[19:06:08] <rayh> Not its length.
[19:08:25] <jlmjvm> thought you had it there for a sec
[19:09:22] <rayh> The idea of getting these sorts of variable values into a gcode file is hard for me to think about.
[19:10:45] <rayh> You really want to add/subtract tool length from probed position?
[19:12:06] <rayh> Or are you trying to measure a tool's length?
[19:12:34] <jepler> g43.1 (new in 2.2) takes the tool's length, not its slot number. there are g41.1 and g42.1 for cutter radius compensation as well. this should be documented..
[19:13:00] <jlmjvm> basically trying to make touch off work for tool offsets also
[19:14:25] <jlmjvm> would be no variables in the g code,just a standard command like g43h1 in the program
[19:14:34] <rayh> jepler: Wah that solves the problem. Do they also write lengths and diametersw.
[19:15:30] <rayh> I built a tool measuring stand long ago. I just used an inverted probe and wrote the probed value using a script.
[19:18:12] <jlmjvm> anyone have an example of g43.1 in use,cant find anything on wiki
[19:18:16] <jepler> sounds similar to what http://cvs.linuxcnc.org/cvs/emc2/nc_files/tool-length-probe.ngc?rev=1.4;content-type=text%2Fx-cvsweb-markup is doing
[19:19:00] <jepler> that code above has an example ^^
[19:19:08] <jepler> the description of the code is in the documentation .. html version here: http://linuxcnc.org/docs/2.2/html/gcode_main.html#sub:G43,-G49:-Tool
[19:20:58] <jlmjvm> thanks for the link
[19:23:15] <rayh> I only see one g43 in there but that's the idea.
[19:33:04] <jlmjvm> no wonder i couldnt make the g43.1 work,didnt have the [ ]
[19:46:57] <jlmjvm> jepler:is there any way to have a few extra lines for the var for h# values,problem is if i add h1,etc to usr/bin/axis and i try to use it during touch off it saysp value out of range for g10 l2
[19:47:35] <jlmjvm> p value,which i assume means it only setup for so many fixture offsets
[19:49:02] <jlmjvm> want H1 in the touch off panel to store number at #1 in the .var
[19:49:40] <cradek> jlmjvm: touch-off sets coordinate system offsets like g54. it has nothing to do with tool length or g43
[19:49:45] <jlmjvm> then it could be called with g43.1k[#1]
[19:51:18] <jlmjvm> im aware of that,but if it had some extra lines to write to it could be used for tool length touch off also
[19:52:06] <cradek> oh I see what you're trying to do, I think
[19:52:13] <cradek> bbl
[19:52:28] <jlmjvm> later
[19:53:14] <jlmjvm> i just want h1 in my touch off box to write to #1 in my var file
[19:53:36] <jlmjvm> dont know how to redirect
[19:57:35] <skunkworks> I don't exactly understand what you are trying to do. I had set the touch off switch and new the exact location off of the table. I could go and probe the switch with the tool and then set the g10 l whatever by the probled variables. I don't think it workes as slick as setting tool lengths. (something more I need to play with)
[19:59:26] <jlmjvm> skunkworks:im just trying to have a button set the tool length
[19:59:38] <skunkworks> ah
[20:01:15] <jlmjvm> i have a H1 im my touchoff box now,if i could just get that to write to #1 in my var file it would work if you used g43.1 in your programs
[20:01:53] <alex_joni> jlmjvm: that's not a 5 minute hack job
[20:03:34] <jlmjvm> i didnt figure it would be,but i feel that its very important
[20:04:22] <skunkworks> hi alex
[20:04:31] <alex_joni> hi samco
[20:05:38] <jlmjvm> i knew my first day in python would be ruff,lol
[20:05:47] <skunkworks> I got a vextra 5 phase stepping motor and drive out of a film machine.
[20:08:00] <skunkworks> why do I always say vextra.. berger lahr
[20:08:21] <skunkworks> maybe because vextra has lots of 5 phase stepping drivesw
[20:31:54] <jlmjvm> if you use the g43.1 make sure you use the # inside the [ ] or it will move by what ever the number is
[20:32:40] <jlmjvm> #5343 was the var number,forgot the #,moved z-5343.00
[20:33:47] <jlmjvm> do emc,s limits not apply in mdi?
[20:34:19] <user__> user__ is now known as skinnypuppy34
[20:39:01] <alex_joni> jlmjvm: sure they do
[20:40:56] <skinnypuppy34> Limits work fine for me in mdi
[20:41:16] <skinnypuppy34> But still on occasion g2g3 will produce no motion in mdi
[20:41:31] <jlmjvm> type this in mdi g43.1k[5343]
[20:42:27] <jlmjvm> my limits for z is -4.5",that showed z-5343.00
[20:42:45] <jlmjvm> figured i would have got an alarm
[20:44:31] <alex_joni> linear move in MDI would exceed limits
[20:45:35] <jlmjvm> did you get an alarm?
[20:45:39] <alex_joni> yes
[20:45:53] <jlmjvm> im not getting one
[20:46:03] <alex_joni> oh.. sorry.. tried the G1X[5343]
[20:50:58] <skinnypuppy34> alex, Is there anything in particular I can troubleshoot when the g2 g3 won't work in mdi? It's not a real bother since a quick f1 f2 reset will make it work again. But if it's something I can look into for you guys I'll be happy to when it happens next.
[20:53:58] <jlmjvm> alex_joni:is there actually even motion occuring at this point or is it just the gui orienting to the offset and no movement
[20:54:13] <alex_joni> jlmjvm: G43.1 doesn't command movement
[20:54:24] <alex_joni> it will just adjust the tool length, so only GUI
[20:54:34] <jlmjvm> great
[20:54:37] <alex_joni> skinnypuppy34: hmm.. that's strange
[20:54:49] <alex_joni> jlmjvm: but you'll probably get an error as soon as you try to move
[20:55:00] <jlmjvm> yep,i bet so
[20:55:08] <jlmjvm> excellent
[20:55:18] <alex_joni> * alex_joni tries to see if the standard says anything about this
[20:56:11] <jlmjvm> im set now so i have 4 fixture offsets and 5 h numbers available when i use the touch off box
[20:56:42] <jlmjvm> and call the h# with the g43.1
[20:57:17] <jlmjvm> a hack job indeed if ever i saw one
[20:57:34] <jlmjvm> but its doing it
[20:58:28] <alex_joni> jlmjvm: I don't think G43.1K<big number> is a wrong g-code
[20:58:41] <alex_joni> it's just a busted program imo
[20:59:00] <skinnypuppy34> Yeah, g1 g0 moves fine but after repeated g2or g3 < so I can manual knee plunge> g2g3 will produce no movement actual , nor in axis display, but the code will move into mdi history with no error produced as if it thought it ran it. fenn said he had seen something similar and wanted to know if I could reproduce it
[21:00:12] <jlmjvm> if you use g43.1[#5343] it reads that var number,leave out the # and it moves that amount
[21:00:24] <alex_joni> it doesn't move..
[21:00:35] <skinnypuppy34> Doesn't seem to be a problem with gobs of circles in a .ngc
[21:01:06] <alex_joni> skinnypuppy34: you can turn debug on from the AXIS menu, and see if the commands are still sent or not
[21:01:17] <alex_joni> in order to do that you need to run from a terminal
[21:03:00] <jlmjvm> right the gui does,was just suprised when i left the " # " out of g43.1[#5343]
[21:06:01] <skinnypuppy34> alex, thank you . I'll see what I can do with it
[21:06:45] <alex_joni> jlmjvm: you can do the same thing if you enter 5343 as the tool length inside your tool table
[21:06:48] <alex_joni> and G43Hxx
[21:06:57] <alex_joni> so I think that behaviour is correct
[21:10:21] <jlmjvm> for var number 5343 i have a value of -.500,if i do g43.1k[#5343] the gui shows a .500 shift,if i do g43.1k[5343] the gui shows a -5343.00 shift
[21:10:38] <jlmjvm> it was a typo originally
[21:11:17] <alex_joni> jlmjvm: yes, and I think that's the correct behaviour
[21:11:27] <alex_joni> the [] are irrelevant for simple numbers
[21:12:19] <jlmjvm> you dont have to use them
[21:20:59] <jlmjvm> this method for touching tool offsets is working pretty good
[21:22:04] <jlmjvm> just wish there was more room in the var,but its limited to the g10 l2 which only appears to have 9 offsets
[21:22:45] <jlmjvm> my touchoff box has g54-59 and h1-5
[21:22:56] <jlmjvm> g54-57
[21:48:07] <alex_joni> jlmjvm: you can't extend the number of offsets further than 9
[21:48:23] <jlmjvm> how so?
[21:48:40] <alex_joni> however I don't see a reason not to add 5 more fields for tool touch off, instead of using the fixture offsets
[21:49:06] <alex_joni> jlmjvm: they are defined by the standard, are hardcoded in the interpreter, GUI, who knows wherelese
[21:49:32] <[JMS-OPC]> http://www.cricut.com <-- a nifty little cnc machine i saw in an infomercial today at 9 am....
[21:49:51] <[JMS-OPC]> it was on tv, they are selling cnc machines to old ladies now!
[21:50:11] <jlmjvm> my bad alex,i thought you said you could have more than 9
[21:50:35] <alex_joni> [JMS-OPC]: nice :D
[21:52:17] <[JMS-OPC]> i don't see anything about it hooking up to a computer... shouldn't be too hard to hack that together though
[21:53:49] <alex_joni> custom cartridge anyone?
[21:54:19] <[JMS-OPC]> now shipping.. custom cartridges with EMC2 and Linux!
[21:54:23] <[JMS-OPC]> lol...
[21:55:07] <[JMS-OPC]> geez... the cartridges are like 99 bux a pop
[21:59:26] <tomp> maybe it can cut gasket material and become useful (er)
[21:59:57] <[JMS-OPC]> good idea! who needs a carb gasket?
[22:00:09] <skinnypuppy34> g90 g1 x0 y0 f10
[22:00:26] <skinnypuppy34> grr sorry
[22:00:44] <[JMS-OPC]> * [JMS-OPC] goes to zero in inch mode...
[22:02:01] <JymmmEMC> Wonder how hard it would be to hack it?
[22:04:17] <alex_joni> tomp: somehow I suspect it can't
[22:04:28] <alex_joni> it works by placing the paper on some cutting mat
[22:04:42] <alex_joni> and it seems it can't cut through the mat.. only the paper
[22:05:07] <[JMS-OPC]> i'm sure you could always solder new wires onto the motors from your own control... :)
[22:05:23] <[JMS-OPC]> who is going to build the worlds first EMC controlled Cricut machine?
[22:05:35] <[JMS-OPC]> it would be like scrapbooking on turbo juice!
[22:06:10] <tomp> just make your own with the swivel knife suggestions and emc2 ( yeh >just< :)
[22:06:49] <tomp> but include swivel creaser and swivel perferator
[22:06:51] <[JMS-OPC]> just buy the spare knife part from the webbsite and stick it in your collet i guess...
[22:07:33] <[JMS-OPC]> mastercam would make programming it cake!
[22:07:56] <[JMS-OPC]> image import...countour then post
[22:08:11] <anonimasu> you forgot what mastercam costs
[22:08:12] <anonimasu> :p
[22:08:17] <[JMS-OPC]> ohhhhh.............
[22:08:26] <[JMS-OPC]> they charge for it? :|
[22:08:36] <anonimasu> are you serious?
[22:08:47] <anonimasu> like 14000$
[22:09:05] <[JMS-OPC]> rofl... i'm kidding... i know it's expensive... they don't even list prices on the website...
[22:09:05] <tomp> what there's a industry for this ? http://www.aeronautauto.com/ProductList.aspx?CategoryID=4
[22:09:33] <anonimasu> tomp: Yes, they cut all kinds of fabric
[22:09:37] <anonimasu> carbon fibre..
[22:09:42] <anonimasu> stuff for gloves..
[22:09:44] <anonimasu> things like that
[22:09:55] <anonimasu> or vinyl..
[22:10:01] <[JMS-OPC]> you could start making knock of nike's if you had one of those to cut leather!
[22:10:03] <tomp> i cut by hand with an electric wheel knife... used to run a flag shop
[22:10:05] <anonimasu> yep
[22:12:45] <[JMS-OPC]> http://www.aeronautauto.com/images/Products/Tiger_Angel3S.jpg i need one just so i can make myself outfits like this...
[22:14:20] <tomp> used tro do something called '
[22:14:38] <tomp> trapunto, which could be cnc'd
[22:15:16] <tomp> and for the filling we used airguns with yarn, injecting in between the layers, to give the forms height, like bas-relief
[22:18:43] <anonimasu> :)
[22:19:02] <anonimasu> lol
[22:34:37] <alex_joni> good night all
[22:35:45] <a-l-p-h-a> jmkasunich, you around? anything on the mesa? (It's been months since I've asked)
[22:37:53] <jmkasunich> no progress all summer
[22:38:03] <a-l-p-h-a> np
[22:38:04] <a-l-p-h-a> :D
[22:38:08] <jmkasunich> I'm hoping to get back onto it (and my CNC projects in general) this winter
[22:38:09] <a-l-p-h-a> I've been way to busy with work...
[22:38:13] <a-l-p-h-a> cool.
[23:01:33] <Ziegler> http://slashdot.org/article.pl?sid=07/11/18/180249&from=rss
[23:02:25] <davidf> Hi
[23:02:26] <[JMS-OPC]> i reead that a few hours ago... explains a lot about why my head is so screwed up... mom had me stuck to the tv for that show when i was a kid... LOL
[23:03:00] <Ziegler> Maybe thats why I smoke a pipe
[23:03:04] <Ziegler> cookie made me do it
[23:03:16] <[JMS-OPC]> uhh what kinda pipe?
[23:03:18] <[JMS-OPC]> ;)
[23:03:21] <Ziegler> that Adults Only Content as a 5 year old
[23:03:38] <Ziegler> Tobacco... I actually grow it too
[23:03:43] <davidf> anyone know why emc2 accepts this code: O101 while [#1 lt #1] but not this...
[23:03:48] <[JMS-OPC]> awesome, how hard is tobacco to grow?
[23:03:53] <rayh> I remember cookie. Seventh grade I think
[23:04:06] <davidf> O101 while [#1 gt #2]
[23:04:08] <[JMS-OPC]> i have a green thumb and i thought about trying to grow some
[23:04:18] <Ziegler> not hard at all... if you can grow tomatoes... you can grow tobacco
[23:04:36] <Ziegler> keeps the pests off can be a bit of a pain
[23:04:38] <[JMS-OPC]> i can grow anything.. i have a green thumb :)
[23:05:24] <Ziegler> you need a good 80-90 days of good weather to get an abundant crop
[23:05:36] <[JMS-OPC]> could you grow it indoors?
[23:05:47] <Ziegler> yes with proper lighting and soil
[23:05:48] <[JMS-OPC]> you should see the indoor "tomatoes" i grew this year...
[23:06:37] <Ziegler> In fact... I just sold some seeds to a Cuban National who is going to grow them to mix with his new line of cigars
[23:07:13] <Ziegler> oops... meant to say Cuban American
[23:07:23] <[JMS-OPC]> how much do you grow?
[23:07:44] <[JMS-OPC]> my grandparents in kentuck always had about 10 acres or more of tobacco growing when i would go visit....
[23:07:44] <Ziegler> not much... only about 20 plants
[23:07:51] <[JMS-OPC]> kentucky*
[23:08:08] <fenn> davidf: you sure that's what you typed in? seems to work for me
[23:08:11] <Ziegler> I mostly grow it for a bit of personal playing around with, and then sell the seeds
[23:08:23] <[JMS-OPC]> i used to smoke it straight from the barn.. and cough my head off when i was a youngin
[23:08:28] <[JMS-OPC]> lol**
[23:08:31] <Ziegler> gag hehe
[23:08:49] <davidf> fenn, thanks yes, I'm sure. Had this problem before too..
[23:08:57] <fenn> does it give an error message?
[23:09:13] <davidf> yes Bad character used on line xxx
[23:09:35] <davidf> tried ge $ that doesn't work either
[23:09:38] <skunkworks> pastebin your program
[23:09:40] <davidf> no $
[23:09:44] <skunkworks> pastebin.ca
[23:09:46] <Ziegler> might be a previous line
[23:10:13] <davidf> not previous cause when I change it to lt it's fine.
[23:10:24] <davidf> whats the link for pastebin?
[23:10:35] <davidf> thanks.
[23:10:53] <davidf> just a sec...
[23:12:52] <davidf> Here it is... http://www.pastebin.ca/781660
[23:13:44] <[JMS-OPC]> i need to learn about macros in g-codes... i have never dove that deep into g-code programming
[23:13:55] <davidf> shoot. I see it. no brackets on next line.
[23:14:08] <fenn> [JMS-OPC]: better to learn a real programming language i think
[23:14:11] <jepler> there's one mistake on line 51. You have #11=#11-#10. It must be: #11=[#11-#10]
[23:14:28] <jepler> whenever there's an operator, it must be enclosed in square brackets
[23:14:37] <davidf> jepler, right. got it. darn!
[23:14:46] <[JMS-OPC]> i started a "teach yourself python" course.. it'll take me a while to get through it though.. i have to do it in sections
[23:14:53] <davidf> I knew that and looked right at it too.
[23:14:57] <[JMS-OPC]> i got python installed on this windows box...
[23:15:12] <[JMS-OPC]> and i got the EMC2 box with python on it
[23:15:39] <davidf> I did that code with my little program I wrote that converts variables to # vars
[23:15:51] <davidf> python program.
[23:16:19] <fenn> um, why?
[23:16:56] <[JMS-OPC]> so far i have figured out how to do math with python... (add, subtract, divide and multply) i also learned how to format strings and print them to screen... and i started in on the chapter about variables in python
[23:17:05] <[JMS-OPC]> i have a ways to go still but it's soaking in
[23:17:07] <stiles> speaking of macros does emc have support for g65 g66 and g67?
[23:17:22] <fenn> stiles: what are those supposed to do?
[23:17:29] <davidf> Oh, because when I have a program with say 20 variables and 10 loops it gets kind of hairy to write with nothing but #1, #2, #37...
[23:17:59] <fenn> davidf: you can do named variables in gcode, like looks like this: #<local variable> #<_global variable>
[23:18:12] <stiles> g65 is a simple macro call, g66 is a modal macro call and g67 cancles the modal call
[23:18:26] <fenn> no, emc only does subroutines
[23:18:37] <[JMS-OPC]> so you program with "endmill #2" and your python script changes it to.. T2 or something of the sorts?
[23:19:19] <fenn> though i'm not sure exactly what a modal macro call is, or even what a simple macro call is for that matter
[23:20:11] <fenn> davidf: still no named loops though :\
[23:20:14] <davidf> here's the input file for that code... http://www.pastebin.ca/781679
[23:20:22] <stiles> g65 P2000 X Y Z would execute program number 2000 and input the values x y and z into the macro
[23:20:35] <[JMS-OPC]> davidf: i see... neat
[23:20:56] <davidf> fenn, I didn't know that. Looks like I reinvented the wheel, huh?
[23:21:05] <fenn> %s/\$(.*)\$/#\1/g
[23:21:18] <fenn> er,
[23:21:26] <fenn> %s/\$(.*)\$/#<\1>/g i think
[23:21:45] <fenn> ignore me
[23:22:01] <davidf> ??
[23:22:03] <stiles> if you had a macro for a pocket but you needed to make 4 pockets but with just different X coridanates you would use g66
[23:22:44] <Ziegler> m codes?
[23:22:50] <fenn> davidf: open your file in vim, type ':' and then paste that gibberish
[23:22:51] <Ziegler> user defined
[23:23:23] <davidf> fenn, um, what's vim? sorry...
[23:23:37] <davidf> an editor?
[23:23:42] <stiles> VI improved, text editor
[23:23:52] <[JMS-OPC]> what's the g-code for ordering pizza?
[23:23:54] <[JMS-OPC]> lol
[23:24:05] <davidf> all I have on this box is gedit for gnome.
[23:24:26] <[JMS-OPC]> sudo apt-get install vim?
[23:24:41] <davidf> try a 180 degree arc with pepperoni
[23:25:27] <davidf> #<localvariable> pepperoni
[23:25:42] <[JMS-OPC]> good call davidf!
[23:25:48] <fenn> oh yeah, vim wants escaped parentheses, even better
[23:26:19] <[JMS-OPC]> i couldn't handle vim.. gedit is much more my style....
[23:26:56] <davidf> I'll go get vim...
[23:27:37] <fenn> i was trying to show the power of regular expressions, but then it turned around and bit me
[23:27:43] <skinnypuppy34> g00 Dominos g2/g3 g28 ?
[23:29:16] <BigJohnT> if your running unbuntu vim is all ready on there
[23:29:41] <davidf> yeah, just found it...
[23:29:55] <[JMS-OPC]> http://www.clubit.com/product_detail.cfm?itemno=A4842001 <- low cost cnc control motherboard?
[23:30:01] <[JMS-OPC]> maybe???
[23:30:13] <davidf> tried typing that in : %s/\$(.*)\$/#<\1>/g but just got pattern not found...
[23:32:01] <fenn> try this :%s/\$\(.\{-}\)\$/#<\1>/g
[23:32:30] <[JMS-OPC]> tha tlooks like you are typing xusswords and the bad word filter is on, fenn... lol
[23:32:38] <[JMS-OPC]> cusswords*
[23:32:44] <fenn> wusswords
[23:33:09] <fenn> oh it looks like vim ignores /g too
[23:33:53] <fenn> so just do that three times i think
[23:34:25] <fenn> actually this is probably going to fail miserably as soon as emc gets to it, because named variables are local
[23:34:47] <davidf> still get pattern not found.
[23:35:09] <fenn> you opened the file in vim right?
[23:35:47] <davidf> oh, no, I just typed that string in. Gotta hold my hand here I think.
[23:36:00] <davidf> what file?
[23:36:12] <fenn> the g-code you just put on pastebin
[23:36:55] <davidf> ok
[23:37:46] <fenn> you might want to back it up if you havent already
[23:40:10] <davidf> fenn, I'm not having much luck here. I put in the g code and then entered that string at the bottom and hit enter. Nothing happened.
[23:40:18] <fenn> O RLY
[23:40:20] <davidf> What is it supposed to do?
[23:40:43] <fenn> are you sure nothing happened?
[23:41:34] <fenn> its supposed to change things like $this$ to #<this>
[23:42:15] <fenn> but for some reason it only does the first one on the line
[23:44:14] <fenn> though now that i think about it, you should do #<_this>
[23:48:07] <davidf> fenn, I get the idea. vim looks very powerful. I'll check it out when I get time. Right now, I can't get around enough to even figure out the basic keyboard commands.
[23:48:37] <fenn> yeah it takes some time and effort to learn. anyway here's the result http://pastebin.ca/781732
[23:49:26] <rayh> bingo and now you know why rayh got weaned from vi/vim after RedHat5.0
[23:49:29] <fenn> i thought maybe you had more programs like that and would want to know how to convert them
[23:50:31] <fenn> i like kate also, which is quite clicky and follows windows keyboard conventions, and also does regular expressions
[23:50:34] <davidf> I get it. It seems to do exactly what my code converter program does
[23:51:11] <davidf> kate sounds more my speed. (tricycle velocity)
[23:51:23] <davidf> :)
[23:51:46] <fenn> well, it has a gentler learning curve, but i wouldnt call it a tricycle
[23:51:59] <davidf> big wheel?
[23:52:17] <fenn> maybe one of those powerwheel things
[23:52:59] <davidf> ha. What advantages would I have if I used the #<_thingy> ?
[23:53:27] <fenn> http://tbn0.google.com/images?q=tbn:Z-u10wMaL8z_7M:http://www.intersourcetoys.com/shop/images/303%2520(pink).jpg
[23:53:45] <fenn> davidf: well, you dont have to run your converter program after editing it
[23:54:16] <davidf> OK.
[23:54:45] <[JMS-OPC]> THE BARBIE JEEP IS HERE!!! everyone hop in
[23:55:15] <davidf> So like this? #<foo> =1 G00<_foo> (?)
[23:55:44] <fenn> um, no
[23:56:05] <fenn> #<_foo> = 1
[23:56:19] <fenn> G00 X #<_foo>
[23:56:25] <davidf> right just a typo.
[23:56:56] <fenn> i think if you do = and g0 on the same line, it uses the value from the previous line
[23:57:13] <davidf> So that's all I had to do instead of writthing that whole program, huh? Where can I get a good treatise on G cade?
[23:57:41] <davidf> ha. writthing sounds about right. LOL
[23:57:44] <fenn> this stuff is all unique to emc, and i just added the relevant info to the manual yesterday
[23:58:16] <davidf> Oh. yeah, you mentioned you were working it out before. Now I remember.
[23:58:26] <davidf> Good job! :)
[23:58:35] <fenn> eh, i didn't write it
[23:58:39] <fenn> i hate g-code :)
[23:59:09] <davidf> Oh. I remember someone saying they were working on it. ?
[23:59:11] <fenn> http://linuxcnc.org/docs/devel/html/gcode_main.html#r2_3_3
[23:59:20] <fenn> yeah probably either lerman or lawrenceg?
[23:59:59] <davidf> yeah I think it was one of those two. lerman I think, maybe.